This study will test for the presence of the DNA "replication error" phenotype in oral squamous cell carcinoma. The hypothesis to be tested is that DNA replication errors occur in a significant proportion of patients with oral squamous cell carcinoma. Genetic instability has recently been identified as a major pathophysiologic lesion underlying development of a variety of inherited and sporadic cancers. Use of the replication error phenotype assay to identify tumors with genetic instability is an important new advance in cancer research. To date, the prevalence of the replication error phenotype has not been rigorously evaluated for oral/head and neck cancer. Therefore we will determine the prevalence of the replication error phenotyPe as defined by microsatellite instability in single and multiple primary oral squamous cell carcinoma. The institutional resources available to the PI are uniquely suited to this study. The rate of oral/head and neck cancer in North Carolina exceeds the national average. As a regional referral center for oral/head and neck cancer, the Bowman Gray School of Medicine/North Carolina Baptist Hospital (BGSM/NCBH) treated over 6,000 individuals for oral/head and neck cancer from 1975-1995. Complete information regarding clinical findings, patient demographics, and exposure to tobacco and alcohol is available in a computerized tumor registry database. Archived tissue specimens (for normal and cancer tissues) are also available for most of these patients. We propose a retrospective study to test for the presence of the replication error phenotype in 100 patients with primary oral squamous cell carcinoma, including 50 individuals with metachronous primary cancers. Microsatellite instability will be determined by PCR amplification of DNA, gel electrophoresis and autoradiography. We will genotype ten different microsatellite loci for DNA from a cancer specimen and from normal DNA for each individual studied. We will test for a significant association between the presence of the replication error phenotype and oral squamous cell carcinoma. We will also test for a significant association with the occurrence of multiple primary cancers, response to treatment and disease outcome. Identification of a significant relationship between the replication error phenotype and oral squamous cell carcinoma could lead to development of a diagnostic test to help predict disease prognosis.